Is there any classification for the item difficulty and the item description ranges of values in item response theory (IRT) and multidimensional item response theory (MIRT).
According to Baker in "The basics of item response theory"
item discrimination in IRT is classified into the following :
none 0
very low 0.01 - 0.34
Low 0.35 - 0.64
moderate 0.65 - 1.34
High 1.35 - 1.69
Very high > 1.70
Perfect + infinity
According to Baker,Hambleton (Fundamentals of Item Response Theory ), and Hasmy (Compare Unidimensional and Multidimensional Rasch Model for Test with Multidimensional Construct and Items Local Dependence) item difficulty is classified into the following :
very easy above [-2,...]
easy (-0.5,-2)
medium [-0.5,0.5]
hard (0.5,2)
very hard [2,..]
Could the item discrimination and item difficulty classification be also used in MIRT
Dear Mona
these two books considered as a good references at this domain, an acceptable item which has good characters and have a significant item fit index too considered as a good item whatever the scale is unidimensional or multidimensional.
Dear
hope this will help you
regards
http://scholarsarchive.byu.edu/cgi/viewcontent.cgi?article=1223&context=etd
Hi Rana,
Thank you very much !!!
Mohamad, they may depend on the question type but i also believe there are many other question features that could affect the values
Hi Mona,
Would you please share the specific reference (Author, year and page) for your item difficulty classification in IRT?
On discrimination parameters
It depends on the item-fit based on RMSEA values (equal or less than 0.05)
In my case, I deleted items with lowest discrimination and RMSEA value of 0.026.
When it comes to item discrimination, values of around 1 are typical in educational measurement. In health measurement, however, these values are usually much higher: values of around 4 are not uncommon! The latter items are typically also polytomous, whereas items used in education are often dichotomous.
When it comes to the item difficulty parameters, what you find acceptable depends very much on the purpose of your test (of course this goes for the discrimination as well) and the population of interest. If your aim is to set a diagnosis, for example, for which you use a cut-off score, you may be interested in having items with a b-value around that cut-off score and the other ranges are not very important. If you want to adequately measure the entire population distribution of interest, it is important to have a closer look at the "targeting" of your instrument.
- Badges
- Science topic
- Similar topics
- Correction